Alterations in protein and amino acid metabolism in rats fed a branched-chain amino acid- or leucine-enriched diet during postprandial and postabsorptive states

Male Wistar rats (BioTest, Konarovice, CR) were housed in standardised cages in quarters with controlled temperature and a 12-h light–dark cycle. All rats received the standard laboratory diet (SLD) ST-1 (Velas, CR) and drinking water ad libitum. All procedures involving animals were performed according to the guidelines set by the Institutional Animal Care and Use Committee of Charles University. Animal Care and Use Committee of Charles University in Prague, Faculty of Medicine in Hradec Kralove specifically approved this study. L-[3,4,5-³H]phenylalanine was purchased from American Radiolabeled Chemical, Inc. (St. Louis, MO, USA). Chemicals were obtained from Sigma Chemical (St. Louis, MO, USA), Lachema (Brno, CR), Waters (Milford, MA, ISA), Biomol (Hamburg, Germany), and Merck (Darmstadt, Germany).

A total of 120 male Wistar rats weighing approximately 200 g each were randomly divided into three groups fed an SLD or a diet in which 10 % of the basal diet was replaced by a mixture of valine, leucine, and isoleucine in ratios of 1 : 1 : 1 (HVLID, high valine, leucine, and isoleucine diet) or leucine (HLD, high leucine diet). The mixtures were used to prepare pellets with the same physical properties and consistency. Estimated contents of L-valine, L-leucine, and L-isoleucine (g/kg diet) are shown in Table 1. These contents resemble a high-dose supplementation in which adverse outcomes for the monitored variables were not reported [14‐16].

The animals consumed tested diets for 2 months and were sacrificed between 7 and 8 a.m. on the last day of the study protocol. Half of the animals in each group were sacrificed in the fed state, and the other half were sacrificed after an overnight fast. Two separate studies were performed. Tissue protein synthesis rates were measured using the flooding dose method (L-[3,4,5-³H]phenylalanine) in the first study. Alterations in amino acid concentrations in body fluids and various parameters of protein and amino acid metabolism were estimated in the second study.

The rats were injected intravenously with a flooding dose of L-[3,4,5-³H]phenylalanine (50 μCi/100 g b.w.) combined with unlabelled L-phenylalanine (150 μmol/100 g b.w.) 10 min before the sacrifice by exsanguination via the abdominal aorta [17]. Small pieces (approximately 0.1 g) of soleus (SOL), gastrocnemius (GM), and extensor digitorum longus (EDL) muscles, liver, kidneys, and jejunum were quickly removed and frozen in liquid nitrogen. Tissue samples were homogenized in 6 % (v/v) perchloric acid, and the precipitated proteins were collected via centrifugation for 5 min at 12,000 g. The supernatant was used for the measurement of L-[3,4,5-³H]phenylalanine specific activity. The pellet was washed three times and hydrolysed in 2 N NaOH. Aliquots were taken for protein content [18] and radioactivity measurements. The fractional rate of protein synthesis (FRPS) was calculated according the formula derived by McNurlan et al. [19]:

where S _b and S _a are the specific activities (dpm/nanomole) of protein-bound phenylalanine and tissue-free phenylalanine in the acid-soluble fraction of tissue homogenates, respectively, and t is the time (days) between isotope injection and tissue immersion into liquid nitrogen. The value of 274 μmol phenylalanine/g protein was used for the calculation of protein-bound phenylalanine specific activity [20]. Sample radioactivity was measured using a liquid scintillation radioactivity counter LS 6000 (Beckman Instruments, Fullerton, CA, USA).

Amino acid concentrations were determined in the supernatants of deproteinised blood plasma and tissue samples using high-performance liquid chromatography (Aliance 2695, Waters, Milford, MA, USA) after derivatisation with 6-aminoquinolyl-N-hydroxysuccinimidyl carbamate. The intracellular concentration of each amino acid was calculated by subtracting the free extracellular portion from the total amount, assuming the plasma concentration to be equal to the concentration in the interstitial fluid as described by Bergstrőm et al. [21]. Total tissue water was measured from the tissue weight obtained after drying for 24 h at 90 °C. The determination of extra- and intracellular water was based on the chloride method according to Graham et al. [22]. BCKA concentrations in blood plasma were measured using liquid chromatograph (Shimadzu, Kyoto, Japan) after precolumn derivatisation with o-phenylenediamine [23].

The chymotrypsin-like activity of proteasomes was determined using the fluorogenic substrate Suc-LLVY-MCA [24] as follows. The muscles were homogenised in 0.4 ml of ice-cold 20 mM Tris buffer, pH 7.5, containing 2 mM ATP, 5 mM MgCl₂ and 1 mM dithiothreitol. The homogenates were centrifuged for 10 min at 18,000 g at 4 °C. Cellular supernatants (0.1 ml) were incubated with 0.1 ml of substrate Suc-LLVY-MCA (0.1 mM) with or without inhibitor MG132 (0.02 mM) for 1 h on ice. A volume of 0.4 ml of 100 mM sodium acetate buffer (pH 4.3) was added to stop the reaction. Sample fluorescence was immediately determined at an excitation wavelength of 340 nm and emission wavelength of 440 nm (Tecan Infinite^TM 200). The standard curve was established for 7-amino-4-methylcoumarin (AMC), which permitted the expression of CHTLA as nmol of AMC/g protein/hour. The activity was adjusted for the protein concentration of the supernatant. Differences after the subtraction of inhibited from non-inhibited activities were used for calculations.

The activities of cathepsin B and L were determined using the fluorogenic substrate Z-FA-MCA [25, 26] as follows. Tissue samples (approximately 20 mg) were homogenised in 0.6 ml of ice-cold 300 mM sodium acetate buffer, pH 5.0, containing 4 mM EDTA, 8 mM dithiothreitol and 0.2 % Triton X-100 (v/v). The homogenates were allowed to stand for 30 min on ice and centrifuged for 30 min at 18,000 g at 4 °C. Cellular supernatant (0.01 ml) were incubated with 0.19 ml of substrate Z-FA-MCA (0.1 mM) with or without the inhibitor Z-FF-FMK (0.04 mM) for 30 min at 37 °C. The reaction was stopped by the addition of 1 ml of 100 mM sodium acetate buffer, pH 4.3, and the activities of cathepsin B and L were determined as described above for CHTLA.

Glutamine synthetase activity was determined using a colorimetric assay based on the formation of γ-glutamyl-hydroxamate [27]. The catalytic concentration of alanine aminotransferase was determined from the rate of decrease of NADH, measured at 340 nm, by lactate dehydrogenase [28]. Enzyme activities are reported per milligram of protein of reaction mixture. Plasma levels of urea, creatinine, ALT, AST, glucose, triglycerides, and cholesterol were measured using commercial tests (Boehringer, Mannheim, Germany; Elitech, Sées, France and Lachema, Brno, CR).

The intake of HVLID was lower than the intake of SLD and HLD in the initial phase of the study. Differences in daily food intake were not significant beginning the second week. There were no differences in body weight gain between animals receiving the various diets (Fig. 2). Higher blood plasma concentrations of urea were observed in animals that chronically consumed HVLID or HLD. In HLD fed animals we observed higher ALT activity and lower concentration of LDL cholesterol and atherogenicity index than in controls (Table 2).

Plasma concentrations of BCAA, BCKA, alanine, and glutamine increased significantly in animals fed HVLID. Enhanced concentrations of leucine, KIC, glutamine and alanine were also found in animals fed HLD, but concentrations of valine, isoleucine, KIV and KMV were significantly lower than controls. Overnight starvation decreased the plasma concentration of most amino acids in controls and normalised alterations induced by the intake of HVLID or HLD (Table 3 and Fig. 3).

HVLID increased levels of valine, leucine, and isoleucine and HLD increased leucine and decreased valine and isoleucine concentrations in all muscle types (Table 4). In other tissues the effects of HVLID and HLD were less significant (Table 5). Alanine concentrations were higher in all muscle types of animals that consumed HVLID or HLD and in the jejunum and kidneys of animals that consumed HLD. Glutamine concentrations increased significantly only in muscles of the HLD group. Most of the alterations in intracellular amino acid concentrations that were induced by the chronic intake of HVLID or HLD in the postprandial state disappeared after an overnight fast.

The effect of chronic HVLID or HLD consumption on glutamine synthetase and alanine aminotransferase in muscles was not significant. Overnight starvation had no effect on glutamine synthetase, but a significant increase in alanine aminotransferase activity in muscles was found in the SOL and MG of animals fed various diets and the EDL of animals fed SLD before an overnight fast (Figs. 4 and 5).

The fractional rate of protein synthesis (Fig. 6) was higher in the postprandial state in MG and SOL of animals fed an HVLID and in the jejunum of animals fed an HLD. A decrease in protein synthesis was found in muscles after an overnight fast. The decrease was more pronounced in gastrocnemius muscles of animals that consumed HVLID and in EDL muscles of animals that consumed HLD. Higher values of protein synthesis were observed in the postabsorptive state in the kidneys of HVLID and HLD groups compared to controls.

Consumption of HVLID decreased CHTLA activities in the muscles; HLD decreased CHTLA activities in the muscles, liver, jejunum, and kidneys (Fig. 7). Cathepsin B and L activities (Table 6) increased in the jejunum of the HVLID group. Lower activities were found in the SOL and jejunum of animals that consumed HLD compared to animals fed by HVLID. Effect of HVLID and HLD on CHTLA in muscles in postabsorptive state was insignificant. Cathepsin B and L activities in the jejunum and EDL of HVLID and HLD groups were lower than controls fed SLD before an overnight fast.

The effect of HVLID on muscle weight and protein content was not significant, but lower weight and protein content was found in EDL of animals that consumed HLD. Chronic intake of HVLID and HLD significantly increased the weight and protein content of the kidneys. We also observed higher liver weights in HLD fed animals (Table 7).

It is well established that the rate of BCAA degradation in skeletal muscle is highly responsive to changes in dietary intake. The K_m of BCAA aminotransferases for BCAAs is two- to four-fold higher than tissue BCAA concentrations [29]. Therefore, the rate of transamination leading to the production of glutamate and BCKA responds rapidly to changes in BCAA level. BCKA dehydrogenase activity in skeletal muscle is low, and most BCKA that is produced in muscle is released to the circulation and utilised in other tissues, especially the liver. The rise in plasma levels of all three BCKAs and ketoisocaproate in animals fed HVLID and HLD, respectively, indicates an enhanced load of these keto-acids in visceral tissues with unknown consequences. Enhanced KIC production from leucine may activate alternative pathways of KIC (ketoleucine) catabolism by the cytosolic enzyme KIC-dioxygenase in the liver. The result is an enhanced synthesis of ß-hydroxy-ß-methylbutyrate, which may be involved in the observed decrease in LDL cholesterol and muscle protein breakdown in animals fed HLD [30, 31].

To determine whether adaptive changes in enzyme activities in muscle tissue were involved in the rise in alanine and glutamine in blood plasma, alanine aminotransferase and glutamine synthetase activities were measured in muscles. Although marked alterations have been reported in various conditions, such dexamethasone treatment [27], chronic consumption of HVLID and HLD was without notable adaptive changes in these enzymes. Therefore, the main mechanism leading to enhanced alanine and glutamine production in skeletal muscle should be an increased flux of glutamate through glutamine synthetase and alanine aminotransferase resulting from enhanced glutamate production by BCAA aminotransferase.

A chronically enhanced release of BCKA, alanine, and glutamine from muscle may be related to the observed hypertrophy of the kidneys. Enhanced glutamine load may activate ammonia production in various tissues, especially gut and kidneys, and in the case of impaired ammonia detoxification to urea in liver disease induce symptoms of hepatic encephalopathy [32]. Enhanced ammonia concentrations in blood after BCAA supplementation were found during exercise and following a leucine intake >500 mg⋅kg^-1⋅d^-1 [33, 34]. Nevertheless, alterations in aminoacidemia induced by BCAA intake may also have favourable effects. The example might be positive influence of glutamine on the immune system, protein balance, and gut integrity.

The drop in concentrations of valine and isoleucine in blood plasma and muscles in animals fed high amounts of leucine is likely due to the well-known phenomenon that is referred to as BCAA antagonism [35]. Leucine stimulates the BCKA dehydrogenase complex that controls the rate limiting step in BCAA catabolism leading to enhanced catabolism of all three BCAAs. Depletion of valine and isoleucine pools may also be due to their competition with leucine for transport via the L-carrier system.

Higher protein synthesis rates and decreased CHTLA in the postprandial state in the muscles of animals fed an HVLID are consistent with most in vitro studies and/or studies that assessed the immediate response to BCAA administration [2, 36‐38]. Therefore, our results confirm the regulatory effects of BCAAs on protein turnover, which may exert a positive influence on muscle protein balance. The marked decrease in CHTLA in muscles of animals fed an HLD is consistent with studies that concluded that leucine administration may specifically induce a reduction in protein breakdown without increasing protein synthesis [39‐41]. However, we failed to demonstrate a positive effect of BCAA on muscle protein balance and a negative effect in EDL muscles exerted chronic consumption of excessive amounts of leucine. This was indicated by insignificant changes in weight and protein content in muscles of animals consuming HVLID and the decrease in EDL muscles in the HLD group.

We suppose that the discrepancy between the positive effects of HVLID on protein synthesis and proteolysis and HLD on proteolysis and the insignificant changes in muscle protein content may be explained by metabolic alterations in the postabsorptive state that were induced in our study by an overnight fast. Decreased rates of protein synthesis and CHTLA activities in overnight-fasted animals fed a normal diet represent metabolic adaptation that spares muscle protein in starvation. The more pronounced decrease in protein synthesis gastrocnemius muscles in HVLID group and in EDL muscles in HLD group of overnight-starved animals indicate an impaired metabolic response that is clearly not beneficial for muscle protein balance.

The adverse effect of a leucine-enriched diet is likely related to decreased intracellular pools of valine and isoleucine and preferential leucine oxidation. Both valine and isoleucine catabolites may be utilised in the citric cycle and promote consumption of pyruvate from glycolysis while the end product of leucine is acetyl Co-A, which blocks the entry of pyruvate into the citric cycle. Therefore, preferential leucine oxidation may exert negative effect on glycolysis, ATP production, and muscle performance, and lead to elevations in alanine aminotransferase activities in blood plasma. Enhanced plasma alanine aminotransferase activity without alterations in other markers of hepatocellular damage was reported after dietary leucine excess in muscle disease patients and apparently healthy individuals [42‐44].

The finding of more significant alterations in EDL muscles compared to SOL or gastrocnemius muscles presents additional evidence that muscles composed mostly of white, fast-twitch fibres are more sensitive to various physiological or pathological signals than muscles composed mostly of red, slow-twitch fibres [45, 46].

A number of variables can modify the response to starvation, e.g., duration of starvation, muscle type, and ageing. Combaret et al. [47] reported that the rates of proteasome-dependent proteolysis were 1.5–2 fold higher in muscles in postabsorptive state and that with aging tended to decrease in the postabsorptive state and increase in the postprandial state. A role may play alterations in insulin sensitivity, the BCKA dehydrogenase activity, and amino acid concentrations [35, 48]. It should be noted that only protein synthesis rates can be measured directly under in vivo conditions. Proteolysis can be estimated indirectly, e.g. by changes MuRF1 and MAFbx expression, rates of ubiquitination, and activities of enzymes of ubiquitin-proteasome system [49]. Therefore, the results of the present study cannot exclude significant effects of chronic consumption of HVLID or HLD on proteolysis in postabsorptive state in other experimental conditions.